'Brawl Stars' was conceived as an ambitious mobile game: a multiplayer shooter that should be able to reach a large global audience. With controls much more demanding than those in previous Supercell games, a lot of design and prototyping effort was put into making them as simple and as well-suited for mobile touchscreen devices as possible. At the same time, a very small team with only a handful of developers needed to design a long-term player progression system that would be a good match for the team's limited content creation capacity.After the game soft launched in June 2017, its player metrics weren't quite reaching the ambitious goals. This prompted the game team to try to improve the game: not just polishing it up, but making major changes as well, while that was still possible. What followed was a more than one-year long beta where the team tried to decide between two control models but ended up with a third, and redesigned the progression system twice, both fundamental core aspects of the game. This session will present the different controls and progression solutions, and based on metrics, argue that they were all good designs and any one of these six iterations of core elements could have worked.